ing early life stages when individuals are small, vul- ... life stages of this or any species of lobster. .... taining 2 lobsters of a simdar size as the test lobster, but.
Vol. 188: 179-191.1999
MARINE ECOLOGY PROGRESS SERIES Mar Ecol Prog Ser
Published November 3
The cause and consequence of ontogenetic changes in social aggregation in New Zealand spiny lobsters Mark J. Butler IV'v*,Alistair B. MacDiarmid2, John D. Booth2 'Department of Biological Sciences, Old Dominion University, Norfolk, Virginia 23529-0266, USA ' ~ a t i o n a lInstitute for Water and Atmospheric Research, PO Box 14-901. Kilbirnie, Wellington, New Zealand
ABSTRACT: Ontogenetic changes in the behavior, spatial distribution, or habitat use of a species are presumably adaptations to ecological forces that dlffer in their effect on various life stages. The New Zealand rock lobster Jasus edwardsii is one of several species of spiny lobster that exhibits dramatic ontogenetic shifts in sociality and spatia.1distribution, and we tested whether such changes are adaptive. We first surveyed several natural populations of J. edwardsii to document size-speclfic differences in aggregation. To determine if chemical cues discharged by conspecifics promote aggregation of certain ontogenetic stages, we tested the responsiveness of lobsters of 3 ontogenetic stages (early benth~c juvenile, juvenile, and subadult) to the chemical cues produced by conspecifics of different sizes. Finally, we tethered lobsters of different ontogenetic stages alone and in groups to test the effect of lobster size and aggregation on mortality. Our results offer compelling evidence that pre-reproductive J. edwardsii undergo an ontogenetic change in sociality that alters their spatial distribution and sunrival. Our field surveys show that J. edwardsii are solitary as early benthic juveniles and become social and aggregate as they grow larger. We then demonstrate, using laboratory experiments, that there is a sizespecific increase in the response of pre-reproductive J. edwardsii to the chemical cues of larger conspecifics which facilitates these ontogenet~cchanges in aggregation. Finally, our tethering results confirm that this change in social condition is selectively advantageous: aggregation does not increase the survival of small lobsters, but larger lobsters survive better in groups. Thus, in this study we demonstrate the linkage between ontogenetic changes in the spatial distribution of a species, the behavioral process that creates the pattern, and the selective advantage conferred by these developmental changes.
KEY WORDS: Spiny lobster . Rock lobster . Jasus edwards~i. Ontogeny . Sociality . Aggregation . Predation
INTRODUCTION
Social aggregation is widespread among marine animals. It occurs in mammals and fish of many kinds (Pitcher 1992),and in a diversity of invertebrate groups including: squid, krill, molluscs, (Catterall & Poiner 1983, Stoner & Ray 1993), decapod crustaceans (Atema & Cobb 1980, among others), and others (Ritz 1994). Social aggregations develop because of mutual attraction among conspecifics and the evolution of cooperative group behavior that enhances individual defense, foraging, movement, or reproduction (Wilson 1975).
O Inter-Research 1999 Resale of full arhcle not permitted
Social groups differ from other aggregations that arise indirectly in response to patchy resources. This is the case, for example, where settlement on limited substrates or foraging on concentrated patches of food results in clusters of animals. Social aggregation can be beneficial because it can increase vigilance and group defense against predators, thus reducing the per capita probability of mortality from predation, and possibly improving detection of food resources (see reviews by Bertram 1978, Pitcher et al. 1982, Pulliam & Caraco 1984, Lima & Dill 1990). However, the benefits of aggregation may be counter-balanced by increases in intraspecific competition for food resources (Milinski & Parker 1991), or by increased mortality caused by
180
Mar Ecol Prog Ser 188: 179-191, 1999
predators that also forage in groups and thus locate clumped prey more easily (Major 1978). Aggregation can also be deleterious for cryptic or camouflaged animals if it makes them more conspicuous (Harvey & Greenwood 1978, Owen 1980, Dukas & Clark 1995, Butler et al. 1997).Moreover, the ecological conditions favoring social aggregation are not static and vary with changes in predator density, resource (food or habitat) availability, and individual size or developmental stage (i.e. ontogeny; Wilson 1975, Trivers 1985). Ontogenetic changes in social aggregation often take 1 of 2 forms. Social aggregations are common during early life stages when individuals are small, vulnerable, and inexperienced. Aggregation often diminishes among subadults, but can reappear at adulthood for reproductive purposes. Less common are species that are solitary as juveniles and aggregated as prereproductive adolescents or subadults. This type of ontogenetic behavioral pattern can develop, for example, where aggregation by defenseless juveniles increases their detection by predators and thus their probability of mortality (Tinbergen et al. 1967, Treisman 1975, Dukas & Clark 1995, Butler et al. 1997). Cryptic behavior and camouflaged coloration are beneficial to juveniles under these conditions. Spiny lobsters (Crustacea; Palinuridae) are 1 group of marine organisms that exhibit this pattern; most species appear to be solitary when young and then aggregate as they grow larger (Herrnkind et al. 1994, Lipcius & Cobb 1994, Phillips & Booth 1994). Spiny lobsters have complex life cycles involving several ontogenetic stages and habitats. Their larvae are typically planktonic for 6 to 18 mo and then metamorphose into a non-feeding puerulus postlarval stage lasting several weeks. Postlarvae swim and are advected nearshore where they settle in shallow, architecturally complex benthic habitats (Herrnkind et al. 1994).There are no data from which it can be discerned whether spiny lobster larvae or postlarvae actively aggregate or are social, although postlarvae periodically settle in dense aggregations in natural habitats (Booth 1979, Jernakoff 1990, Norman et al. 1994) and on artificial collectors (Phlllips & Booth 1994). Newly settled early benthic juvenile (EBJ; =postpuerulus; see Lavalli & Lawton 1996) lobsters live in holes or dense vegetation, are solitary, and at least 1species (Caribbean spiny lobster Panulirus argus) is non-aggregative at this stage (Herrnkind et al. 1994, Childress & Herrnkind 1996).As they grow through adulthood, lobsters inhabit increasingly larger crevices, which for some species results in a marked shift in their choice of habltat and pattern of aggregation (Berrill 1975, Cobb 1981, Herrnkind & Butler 1986, Herrnkind & Lipcius 1989, Trendall & Bell 1989, Eggleston et al. 1990, Glaholt 1990). Yet, only 1 recent study (Ratchford & Eggleston 1998) has linked
these ontogenetic changes in aggregation with the proximate processes that promote them, and no study has offered evidence that such changes influence survival or fecundity. Findings from several unrelated studies suggest that there are distlnct ontogenetic differences In aggregation of Jasus edwardsii, the New Zealand rock or spiny lobster. All stages are found in rocky habitats; juveniles and adults tend to aggregate in particular dens (MacDiarmid 1991, 1994, MacDiarmid et al. 1991), whereas EBJ occupy small shelters individually (Kensler 1966, Booth 1979, Booth & Bowring 1988, Booth et al. 1991). This increase in the patchiness of lobsters with size may be due to: (1) ontogenetic changes in behavior that enhance social aggregation of larger lobsters, (2) differential mortality of lobsters among shelter locations that creates patches of survivors, or (3) differences in the spatial distribution of crevices of different sizes. If the first hypothesis is true and J. edwardsii actively congregate as large juveniles and adults, then they must have some means of locating one another. Chemical (odor) detection is a likely mechanism; it is used by the adults and subadults of other species, notably Panulirus interruptus (ZimmerFaust et al. 1985, Zimmer-Faust & Spanier 1987) and P. argus (Ratchford & Eggleston 1998). However, it is not known whether J. edwardsii responds to chemical cues, or if the use of chemical cues varies among the life stages of this or any species of lobster. Thus, in this study we sought to: (1) determine whether the aggregative behavior of Jasus edwardsii varied among ontogenetic stages, (2) investigate behavioral mechanisms that might create changes in aggregation, and (3) examine the potential adaptive value of ontogenetic changes in aggregation. We first characterized the natural spatial distribution of non-reproductive lobsters of 3 natural ontogenetic stages: EBJ (
